Scaling Experimentation: from 50 to 300 Tests per Year

Running more tests doesn’t guarantee more growth. At TomTom, I turned “spaghetti testing” into a scalable experimentation process, growing from 50 to 300 experiments per year while aligning every test with real business impact.

Challenge

TomTom’s experimentation efforts were siloed and inconsistent. Different teams used different tools, methodologies, and metrics, leading to duplicated work and unclear insights. Leadership wanted a unified, scalable process that could deliver faster learning and consistent business impact.

Results

  • 50 → 300 tests/year, 5× increase in velocity

  • 33% win rate sustained despite higher volume

  • Experiment participation doubled across teams

  • Standardized experimentation processes across all squads

35%

Improved onboarding process

25%

Increase in user retention

84%

Increase in time spent on website

Process

Audit & Alignment: Mapped out the existing experimentation landscape, identifying bottlenecks and missed opportunities.

Framework Creation: Introduced prioritization frameworks (ICE, PIE, and later a custom AI-powered variant) to focus on high-impact ideas.

Infrastructure: Implemented centralized tracking and shared dashboards in Mixpanel and internal BI tools.

Enablement: Designed internal training sessions to empower non-analysts and PMs to run experiments independently.

Knowledge Sharing: Built a shared database of experiment learnings (Airtable), accessible to the entire organization.

Conclusion

Scaling experimentation was about more than testing volume — it was about transforming culture. By giving teams autonomy, structure, and clarity, we created an environment where experimentation became effortless, efficient, and embedded in the company’s DNA.